An image capture device obtains first image data from a front facing camera on a front face of the device including a display. The first image data represents a subject within a field of view of the image capture device. Facial recognition is performed in response to the first image data from the front facing camera. The front image data is used for determining whether any subject is a priority subject, of whom at least one priority subject image-indicative data is accessible by the image capture device. A region of interest is selected, corresponding to the subject determined to be a priority subject. Automatic focus, automatic exposure, or automatic white balance is performed using the selected region of interest. Second image data from the front facing camera are captured based on the automatic focus, automatic exposure, or automatic white balance using the selected region of interest.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
2. The method of claim 1, wherein performing the facial recognition comprises detecting a face within the first image data.
5. The method of claim 4, wherein the neural network is trained to identify members of a predetermined set of priority subjects comprising the particular subject.
A system and method for training a neural network to prioritize specific subjects in image or video analysis. The technology addresses the challenge of efficiently identifying and prioritizing key subjects within large datasets, such as surveillance footage, medical imaging, or autonomous vehicle perception, where processing all elements is computationally expensive. The neural network is trained to recognize and prioritize members of a predefined set of priority subjects, including a particular subject of interest. This involves training the network to distinguish between relevant and irrelevant subjects, ensuring that processing resources are focused on the most critical elements. The network may use features such as appearance, behavior, or contextual cues to classify subjects and assign priority levels. The system may also incorporate adaptive learning to refine priority criteria over time based on feedback or changing conditions. By focusing on priority subjects, the method improves efficiency and accuracy in applications requiring real-time or high-throughput analysis.
6. The method of claim 1, wherein processing the second image data includes capturing the second image data.
This invention relates to image processing systems, specifically methods for capturing and processing image data to enhance visual information. The technology addresses the challenge of obtaining high-quality image data under varying conditions, such as low light or motion blur, by dynamically adjusting capture parameters or combining multiple image inputs. The method involves capturing a second set of image data, which may be derived from a different sensor, a different time, or a different perspective than an initial set of image data. This second image data is processed to improve the overall image quality, resolution, or clarity. Processing may include techniques such as noise reduction, alignment, fusion, or enhancement to correct distortions or artifacts present in the original image. The method ensures that the final output is optimized for visual analysis, display, or further computational tasks. The invention may be applied in various fields, including surveillance, medical imaging, autonomous vehicles, and consumer electronics, where reliable and high-fidelity image data is critical. By dynamically capturing and processing image data, the system adapts to environmental changes and improves the accuracy and usability of the visual information.
7. The method of claim 6, wherein the second image data is captured using at least one of an automatic focus, automatic exposure, or automatic white balance setting determined based, at least in part, on the selected region of interest.
This invention relates to image capture systems that automatically adjust camera settings based on a selected region of interest (ROI) within a scene. The problem addressed is the need for improved image quality in specific areas of a captured image, particularly when the entire scene contains varying lighting conditions or complex details. Traditional methods often apply uniform settings across the entire image, which may not optimize the ROI for clarity or visual appeal. The method involves capturing a first image of a scene using initial camera settings, such as focus, exposure, or white balance. A region of interest is then selected within this image, either manually or automatically. Based on the characteristics of this ROI, such as brightness, contrast, or color, the system determines optimized settings for at least one of automatic focus, automatic exposure, or automatic white balance. A second image is then captured using these adjusted settings, ensuring the ROI is captured with improved quality while the rest of the scene may be less optimized. This approach allows for dynamic adaptation to varying conditions, enhancing the visibility and detail of the selected area without requiring manual intervention. The method is particularly useful in applications like surveillance, medical imaging, or photography where specific regions demand higher precision.
9. The method of claim 1, wherein the region of interest includes each of the one or more subjects.
This invention relates to a method for analyzing regions of interest in visual data, particularly for tracking or identifying subjects within a captured image or video. The method addresses the challenge of accurately defining and processing areas containing subjects, such as people or objects, to improve detection, recognition, or monitoring in applications like surveillance, security, or automated analysis. The method involves selecting a region of interest that encompasses all subjects present in the visual data. This region is dynamically adjusted based on the subjects' movements or positions, ensuring that the entire area containing the subjects is consistently captured and analyzed. The method may include preprocessing steps to enhance image quality, such as noise reduction or contrast adjustment, before identifying and isolating the subjects within the region. Advanced techniques like machine learning or computer vision algorithms may be used to detect and track the subjects over time, enabling real-time or post-processing analysis. The method ensures that no subject is excluded from the analysis, improving accuracy in applications requiring full coverage of all subjects, such as crowd monitoring or behavioral analysis. The dynamic adjustment of the region of interest allows for efficient processing, as only the relevant areas are analyzed, reducing computational overhead. This approach is particularly useful in scenarios where subjects move unpredictably or where multiple subjects must be tracked simultaneously.
10. The method of claim 1, wherein selecting the region of interest comprises adjusting the region of interest to include the particular subject.
11. The method of claim 10, wherein the selecting further comprises adjusting the region of interest to exclude at least one of the one or more subjects.
13. The image capture device of claim 12, wherein the at least one processor is further configured to detect a face within the first image data.
This invention relates to image capture devices with enhanced face detection capabilities. The device includes an image sensor for capturing first image data and at least one processor. The processor is configured to detect a face within the first image data. The device may also include a display for presenting the first image data and a user interface for receiving user input. The processor may further process the first image data to generate second image data, which may be displayed or stored. The face detection functionality allows the device to identify and analyze facial features within captured images, enabling applications such as facial recognition, portrait mode adjustments, or user authentication. The device may operate in various modes, including still image capture, video recording, or continuous preview modes, and may adjust settings like focus, exposure, or white balance based on detected faces. The face detection may be performed using machine learning algorithms or traditional computer vision techniques, ensuring accurate and efficient identification of faces in different lighting conditions and orientations. This technology is particularly useful in smartphones, digital cameras, and surveillance systems where real-time face detection is required.
16. The image capture device of claim 12, wherein the neural network is trained to identify members of a predetermined set of priority subjects comprising the particular subject.
17. The image capture device of claim 12, wherein the at least one processor is further configured to capture the second image data.
18. The image capture device of claim 17, wherein the at least one processor is further configured to capture the second image data using at least one of an automatic focus, automatic exposure, or automatic white balance setting determined based, at least in part, on the selected region of interest.
20. The image capture device of claim 12, wherein the region of interest includes each of the one or more subjects.
This invention relates to image capture devices designed to improve subject tracking and framing in photography. The problem addressed is ensuring that all subjects within a scene are properly included in the captured image, even when the subjects move or the device is handheld. The device includes a sensor system that detects and tracks one or more subjects in a field of view. A processing unit analyzes the detected subjects to determine a region of interest that encompasses all subjects, ensuring they remain within the frame. The device then adjusts capture parameters, such as zoom or focus, to maintain the region of interest within the image boundaries. The system may also use predictive algorithms to anticipate subject movement and pre-adjust settings accordingly. Additionally, the device may include stabilization mechanisms to compensate for handheld motion, further improving framing accuracy. The invention ensures that all subjects are consistently included in the final image, even in dynamic or unstable shooting conditions.
21. The image capture device of claim 12, wherein the at least one processor is further configured to adjust the region of interest to include the particular subject.
22. The image capture of claim 21, wherein the at least one processor is further configured to adjust the region of interest to exclude at least one of the one or more subjects.
24. The non-transitory computer-readable storage medium of claim 23 wherein processing the second image data comprises capturing the second image data using at least one of an automatic focus, automatic exposure, or automatic white balance setting determined based, at least in part, on the selected region of interest.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
April 8, 2021
October 11, 2022
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.